6\QWKHVLVDQGFU\VWDOJURZWKRI&V)H6HDQHZLURQEDVHGVXSHUFRQGXFWRUZLWK T F b.
This article has been downloaded from IOPscience. Please scroll down to see the full text article.
2011 J. Phys.: Condens. Matter 23 052203 (http://iopscience.iop.org/0953-8984/23/5/052203)
Download details:
IP Address: 192.33.126.162
The article was downloaded on 18/02/2011 at 10:26
Please note that terms and conditions apply.
View the table of contents for this issue, or go to the journal homepage for more Home Search Collections Journals About Contact us My IOPscience
J. Phys.: Condens. Matter23(2011) 052203 (4pp) doi:10.1088/0953-8984/23/5/052203
FAST TRACK COMMUNICATION
Synthesis and crystal growth of
Cs 0.8 (FeSe 0.98 ) 2 : a new iron-based superconductor with T c = 27 K
A Krzton-Maziopa
1,5, Z Shermadini
2, E Pomjakushina
1, V Pomjakushin
3, M Bendele
2,4, A Amato
2, R Khasanov
2, H Luetkens
2and K Conder
11Laboratory for Developments and Methods, Paul Scherrer Institute, CH-5232 Villigen PSI, Switzerland
2Laboratory for Muon Spin Spectroscopy, Paul Scherrer Institute, CH-5232 Villigen PSI, Switzerland
3Laboratory for Neutron Scattering, Paul Scherrer Institute, CH-5232 Villigen PSI, Switzerland
4Physik-Institut der Universit¨at Z¨urich, Winterthurerstrasse 190, CH-8057 Z¨urich, Switzerland
E-mail:kazimierz.conder@psi.ch
Received 16 December 2010, in final form 5 January 2011 Published 19 January 2011
Online atstacks.iop.org/JPhysCM/23/052203 Abstract
We report on the synthesis of large single crystals of a new FeSe layer superconductor Cs0.8(FeSe0.98)2. X-ray powder diffraction, neutron powder diffraction and magnetization measurements have been used to compare the crystal structure and the magnetic properties of Cs0.8(FeSe0.98)2with those of the recently discovered potassium intercalated system KxFe2Se2. The new compound, Cs0.8(FeSe0.98)2, shows a slightly lower superconducting transition temperature (Tc=27.4 K) in comparison to 29.5 in(K0.8(FeSe0.98)2). The volume of the crystal unit cell increases by replacing K by Cs—thecparameter grows from 14.1353(13) to 15.2846(11) ˚A. For the alkali metal intercalated layered compounds known so far,(K0.8Fe2Se2
and Cs0.8(FeSe0.98)2), theTcdependence on the anion height (distance between Fe layers and Se layers) was found to be analogous to those reported for As-containing Fe superconductors and Fe(Se1−xChx), where Ch=Te,S.
(Some figures in this article are in colour only in the electronic version)
1. Introduction
The recent discovery of Fe-based superconductors has triggered a remarkable renewed interest for possible new routes leading to high-temperature superconductivity. As observed in the cuprates, the iron-based superconductors exhibit interplay between magnetism and superconductivity, suggesting the possible occurrence of unconventional superconducting states.
5 On leave from: Warsaw University of Technology, 00-664 Warsaw, Poland.
Other common properties are the layered structure and the low carrier density. Among the iron-based superconductors FeSe1−x has the simplest structure with layers in which Fe cations are tetrahedrally coordinated by Se [1]. The superconducting transition temperature (Tc) of 8 K was found to dramatically increase under pressure reaching a maximum of 37 K, with a rate of dTc/dP ∼ 9.1 K GPa−1—the highest among all the Fe-based superconductors [2]. Additionally, it was found that an applied pressure modifies the electronic phase diagram of FeSe1−x and induces static magnetic order
J. Phys.: Condens. Matter23(2011) 052203 Fast Track Communication
which can coexist with superconductivity [3]. Moreover, the substitution of Te for Se leads to an increase of Tc up to 14 K [4]. Since Te is larger than Se, this effect could not be considered as corresponding to an externally applied pressure.
It was also found that substitution of S for Se increases Tc
slightly [5]. Comprehensive studies of substitutions on the iron site in MxFe1−xSe0.85with the non-transition metals M=Al, Ga, In, Sm, Ba and transition metals M=Ti, V, Cr, Mn, Co, Ni and Cu, were performed by Wu et al [6]. In the first case it was stated that the observed slight change of Tc (within ±2 K) depends on the size of the substituting ions and the doping level, suggesting the importance of the lattice deformation on the superconducting properties. All the systems with transition metals (with the exception of Mn) do not exhibit superconductivity. In the case of M=Cu already 1.5% of doping suppresses superconductivity [7]
and with 10% the compound becomes a Mott insulator [8].
Transition metal doping was also reported in the case of Fe1−xSe0.5Te0.5. Doping with Co (0.05 ! x ! 0.2) and Ni (0.05 ! x ! 0.1) both suppress Tc and lead to a metal–
insulator transition [9]. Doping with Cu (x = 0.05) destroys superconductivity whereas magnetic Mn (x = 0.05) slightly increases Tc[10]. The alkali-metal-doped Na0.1FeSe, where the Na ions are intercalated between FeSe layers, was found to be superconducting withTc=8.3 K [11].
Very recently superconductivity at above 30 K was found in K0.8Fe2Se2 [12]. This is, so far, the highest Tc for Fe- chalcogenides, even though the superconducting fraction is low and the transition is broad. It is also reported [12] that single crystals of several mm3 could be grown from the self-flux.
This new compound is isostructural to layered (122-type) iron pnictides with the space groupI4/mmm[13].
This structure has the same FeSe layers as in (11) FeSe [14], but in (11) FeSe with space groupP4/nmmthese layers are identical with respect to the translation along the z direction (this is obvious, because the unit cell contains only one FeSe layer). In K0.8Fe2Se2 the unit cell is doubled along the z axis. The neighboring FeSe layers along the z direction are shifted by (1/2,1/2,1/2), so that the upwards SeFe4 pyramid is faced with the downwards pyramid along the z direction and the intercalated atom is located between more distant Se atoms along z. The Fe–Fe layer distance in K0.8Fe2Se2increases to 7.0184 ˚A in comparison with 5.5234 ˚A in FeSe [14]. The intercalation of K also increases the Fe–Se bond length within the layers by 2.15%.
In the present work we report on the synthesis and crystal growth of a new analog compound with Cs intercalated between FeSe layers. In comparison with the work of Guoet al [12] we managed by this substitution to significantly increase the superconducting fraction by only slightly diminishing the critical temperature.
2. Experimental details
Single crystals of both potassium and cesium intercalated iron selenides of nominal compositions Cs0.8(FeSe0.98)2 and K0.8(FeSe0.98)2were grown from the melt using the Bridgeman method. Ceramic rods of the iron selenide starting material
were prepared by the solid state reaction technique [14]. The nominal stoichiometry of the starting material that is FeSe0.98
was chosen based on the view of our previous studies [14]
which demonstrated that for this particular Fe/Se ratio the content of secondary phases is the smallest. High purity (at least 99.99%, Alfa) powders of iron and selenium were mixed in an 8 g batch, pressed into rods, sealed in evacuated quartz ampoules and annealed at 700◦C during 15 h. The initially treated material was then ground in an inert atmosphere, pressed again into rods, sealed in evacuated quartz ampoules and thermally treated at 700◦C over 48 h followed by further annealing at 400◦C for another 36 h.
For the single-crystal synthesis a piece of the ceramic rod of FeSe0.98 was sealed in a double-wall evacuated silica ampoule with the pure alkali metals (either potassium or cesium of at least 99.9% purity, Chempur). The quantity of alkali metal used for the synthesis depended on the desired stoichiometry of the final compound. The ampoules were annealed at 1030◦C over 2 h for homogenization. Afterwards the melt was cooled down to 750◦C at the rate of 6◦C h−1and then cooled down to room temperature at the rate 200◦C h−1. Well-formed black crystal rods of 7 mm diameter (diameter of the quartz ampoules) were obtained which could be easily cleaved into plates with flat shiny surfaces.
The Cs0.8(FeSe0.98)2 and K0.8(FeSe0.98)2 crystals were characterized by powder x-ray diffraction (XRD) using a D8 Advance Bruker AXS diffractometer with Cu Kα radiation.
For these measurements a fraction of each crystal was cleaved, powderized, and loaded into the low background airtight specimen holder in an He glove box to protect the powder from oxidation. The K0.8(FeSe0.98)2 polycrystalline sample, which was synthesized the same way as proposed by Guoet al [12], was additionally studied by means of neutron powder diffraction (NPD) at the SINQ spallation source of the Paul Scherrer Institute (PSI, Switzerland) using the high-resolution diffractometer for thermal neutrons, HRPT [15], with the neutron wavelengthsλ=1.494 and 1.886 ˚A. The sample was loaded into a vanadium container with an indium seal in an He glove box. The refinements of the crystal structure parameters were done using the FULLPROF program [16] with the use of its internal tables.
The superconducting transition has been detected by AC susceptibility by using a conventional susceptometer. The sample holder contains a standard coil system with a primary excitation coil (1300 windings, 40 mm long) and two counter- wound pick-up coils (reference and sample coil, each 10 mm long and 430 windings) which are connected to a lock-in amplifier. The frequency used was 144 Hz and the sample holder diameter was 5 mm. Measurements were performed by heating the sample at a rate of 9 K h−1. The susceptometer was calibrated using the superconducting transition of a lead sample showing a 100% superconducting fraction. The raw data were then normalized to the sample volume relative to the one of the Pb calibration specimen.
3. Results and discussion
Room-temperature XRD experiments revealed that the crystals do not contain any impurity phases. The only detected 2
Figure 1.Rietveld refinement pattern (upper—red) and difference plot (lower—black) of the x-ray diffraction data for the crystal with the nominal composition of K0.8(FeSe0.98)2. The rows of ticks show the Bragg peak positions for theI4/mmmphase. The left insert shows the difference Fourier density map atz=1/2 slice obtained from NPD data showing the presence of K at (0.5, 0.5, 0.5), while the colored scale shows scattering density in fm. The right insert shows a picture of the cleaved K0.8(FeSe0.98)2crystal.
Figure 2.Rietveld refinement pattern (upper—red) and difference plot (lower—black) of the x-ray diffraction data for the crystal with the nominal composition of Cs0.8(FeSe0.98)2. The rows of ticks show the Bragg peak positions for theI4/mmmphase. The inset shows a picture of a piece of Cs0.8(FeSe0.98)2crystal.
phase is the tetragonal phase of ThCr2Si2 type (space group I/4mmm). The results of the Rietveld refinement of the XRD patterns are shown in figures1and2for Cs0.8(FeSe0.98)2 and K0.8(FeSe0.98)2, correspondingly. The broad halo on the XRD pattern around 20◦is caused by a sample holder with a plastic dome. This area was excluded from the refinements. For the refinement it was assumed that all Fe and Se sites are fully occupied. The crystallographic data for Cs0.8(FeSe0.98)2 and K0.8(FeSe0.98)2crystals and K0.8(FeSe0.98)2polycrystalline are summarized in table1.
One can note that the atomic displacement parameters (ADP) refined from XRD are quite large in comparison with
Figure 3.Temperature dependence of the AC susceptibility (χ$) for single-crystalline K0.8(FeSe0.98)2and Cs0.8(FeSe0.98)2. The signal has been normalized to a superconducting Pb specimen as described in the text.
Table 1. Structural parameters for Cs0.8(FeSe0.98)2and
K0.8(FeSe0.98)2powderized crystals and K0.8(FeSe0.98)2powder at 290 K obtained from XRD and NPD, correspondingly. Space group I4/mmm(no. 139), Fe in (4d) position (0, 0.5, 0.25); Se in (4e) position (0, 0,z), Cs/K in (2a) position (0, 0, 0). The atomic displacement parameters (represented by the B parameter) for all atoms were constrained to be the same.
Cs0.8(FeSe0.98)2 (XRD) powderized crystal
K0.8(FeSe0.98)2 (XRD) powderized crystal
K0.8(FeSe0.98)2
(NPD) polycrystalline
a( ˚A) 3.9601(2) 3.9092(2) 3.9038(1)
c( ˚A) 15.2846(11) 14.1353(13) 14.1148(6)
Se,z 0.3439(3) 0.3503(3) 0.3560(3)
Cs/K
occupancy 0.771(7) 0.792(10) 0.737(20)
B( ˚A2) 3.37(9) 3.16(9) 1.63(4)
Rp,Rwp,Rexp 4.35, 6.00, 3.09 3.84, 5.10, 2.99 4.91, 6.71, 3.25
χ2 3.77 2.90 4.26
the ones refined from NPD. We believe that this results from a slight degradation of the samples during the XRD measurements due to the non-ideal sealing of the plastic container, inasmuch as the samples are extremely air-sensitive.
It is also supported by the fact that both samples show a pronounced strain-like diffraction peak broadening. The enhanced ADP can lead to systematic errors in determination of the site occupancies. The substitution of Cs for K causes the lattice to expand predominantly in thecdirection, due to the larger ionic radius of Cs (1.78 ˚A) compared to K (1.51 ˚A).
Figure 3 shows the temperature dependence of the AC susceptibility (χ$) for single crystals of K0.8(FeSe0.98)2 and Cs0.8(FeSe0.98)2. The onset of the critical temperature has been determined to be Tc,onset = 27.4 K and 29.5 K for the Cs and K intercalated compounds, respectively. It has to be noted that the diamagnetic signal is larger in the case of the Cs-intercalated FeSe, which might point to a bigger Meissner fraction. A further investigation if this effect is intrinsic to the K1−x(FeSe0.98)2and Cs1−x(FeSe0.98)2families
J. Phys.: Condens. Matter23(2011) 052203 Fast Track Communication
Figure 4.Dependence ofTcon the distance between Fe and the chalcogenide/pnictide layers. The red line shows the dependence as presented by Mizuguchiet al[17] for typical Fe-based
superconductors. The filled circles indicate the data obtained for FeSe under high pressure [2]. The open points indicate the data obtained by this work and our previous reports [18,19].
or if the superconducting fraction sensitively depends on the compositionxis underway.
Intercalation of the alkali metals into the FeSe causes serious structural changes. It was proved by Mizuguchi et al[17] that in iron pnictides and chalcogenides the critical temperature can be correlated with the so-far known ‘anion height’, which is the distance between Fe and chalcogenide (or pnictide) layers in the structure. Figure 4 presents a line (taken from [17]) being a best fit to the experimental data obtained for over 15 differed Fe-based superconducting compounds. The curve shows a relatively sharp peak around 1.38 ˚A with a maximum transition temperatureTc ≈ 55 K (for NdFeAsO0.83). The open symbols in figure 4 depict the anion height to Tc correlation of the samples synthesized in this work and those presented in our previous reports [18,19]. Apparently the newly synthesized compounds with intercalated Cs and K follow very well the universal trend. The tendency in the series FeSe0.98–Cs0.8(FeSe0.98)2– K0.8(FeSe0.98)2is analogous to that reported for high pressure measurements [2] (filled circles in figure4). The steep slope of Tc as a function of anion height suggests that even much higher superconducting transition temperatures might be found in the newly discovered FeSe-based systems by applying either chemical (substitutional) or hydrostatic pressures. Another not yet explored aspect is the relation of magnetism and superconductivity in this system. Anyhow, whether or not superconductivity appears in this system in close proximity to an antiferromagnetically ordered state like in other Fe-based superconductors awaits further investigation. Fortunately, the synthesis method described here is able to provide large single crystals which will allow us to answer the above questions
in detail by applying bulk methods like neutron scattering or muon spin rotation, both at ambient and high pressures.
4. Summary
In conclusion, a new Cs intercalated iron selenide super- conductor(Cs0.8(FeSe0.98)2)was synthesized by the Bridge- man method in the form of large single crystals. The Cs0.8(FeSe0.98)2 compound represents the second member of the alkali metal chalcogenide family. In comparison with the K-analog a larger lattice volume is observed and Tc,onset = 27.4 K is only slightly decreased compared to K0.8Fe2Se2. The large high quality crystals obtained by the method described will allow an in-detail study of fundamental magnetic and su- perconducting properties in this new FeSe family.
Acknowledgments
The authors thank the Sciex-NMSch(Project Code 10.048), the Swiss National Science Foundation and NCCR MaNEP for the support of this study. This study was partly performed at the Swiss neutron spallation SINQ of the Paul Scherrer Institute PSI (Villigen, PSI). We acknowledge the allocation of beam time at the HRPT diffractometer of the Laboratory for Neutron Scattering (PSI, Switzerland).
References
[1] Hsu F Cet al2008Proc. Natl Acad. Sci. USA10514262 [2] Medvedev Set al2009Nat. Mater.8630
[3] Bendele M, Amato A, Conder K, Elender M, Keller H, Klauss H-H, Luetkens H, Pomjakushina E, Raselli A and Khasanov R 2010Phys. Rev. Lett.104087003
[4] Sales B C, Sefat A S, McGuire M A, Jin R Y, Mandrus D and Mozharivskyj Y 2009Phys. Rev.B79094521
[5] Mizuguchi Y, Tomioka F, Tsuda S, Yamaguchi T and Takano Y 2009J. Phys. Soc. Japan78074712 [6] Wu M Ket al2009PhysicaC469340
[7] Williams A J, McQueen T M, Ksenofontov V, Felser C and Cava R J 2009J. Phys.: Condens. Matter21305701 [8] Huang T Wet al2010Phys. Rev.B82104502
[9] Shipra R, Takeya H, Hirata K and Sundaresan A 2009Physica C470528
[10] Zhang A M, Xia T L, Kong L R, Xiao J H and Zhang Q M 2010J. Phys.: Condens. Matter22245701
[11] Liu Z, Fang A, Huang F and Jiang M 2008 arXiv:0808.1784 [12] Guo J, Jin S, Wang G, Wang S, Zhu K, Zhou T, He M and
Chen X 2010Phys. Rev.B82180520(R)
[13] Rotter M, Tegel M and Johrendt D 2008Phys. Rev. Lett.
101107006
[14] Pomjakushina E, Conder K, Pomjakushin V, Bendele M and Khasanov R 2009Phys. Rev.B80024517
[15] Fischer Pet al2000PhysicaB276–278146 [16] Rodr´ıguez-Carvajal J 1993PhysicaB19255
[17] Mizuguchi Y, Hara Y, Deguchi K, Tsuda S, Yamaguchi T, Takeda K, Kotegawa H, Tou H and Takano Y 2010 Supercond. Sci. Technol.23054013
[18] Bendele M, Babkevich P, Katrych S, Gvasaliya S N, Pomjakushina E, Conder K, Roessli B, Boothroyd A T, Khasanov R and Keller H 2010Phys. Rev.B82212504 [19] Bendele Met al2010Phys. Rev.B81224520
4
